kuanhoong.bsky.social
TEDx Speaker, Google GDE, AWS Community Builder, Podcaster, Founder Malaysia R User Group, Artificial Intelligence & Machine Learning Malaysia User Group
326 posts
232 followers
396 following
Prolific Poster
Conversation Starter
comment in response to
post
In contrast to the original Gödel Machine—which demands provable gains for any code changes, a condition that is often impractical—the DGM takes an empirical route: it rewrites parts of its own code and tests the results against coding benchmarks to assess improvement. arxiv.org/abs/2505.22954
comment in response to
post
Special thanks to Google Developer Group (GDG) Cloud KL and the amazing community for the invite and the enthusiastic participation! 🙌
Let’s keep building with AI!
👉 Try NotebookLM here: notebooklm.google
comment in response to
post
🎯 Key takeaways:
- Personalization is the future — your research assistant should know your materials
- AI can reduce research time dramatically while improving insight quality
- Multimodal AI (text + audio) makes information far more accessible and actionable
comment in response to
post
The energy in the room was incredible 🔥 — participants were particularly excited about how NotebookLM transforms information into engaging audio discussions and enables interactive Q&A with their sources.
comment in response to
post
✅ Creating mind maps and structured content like FAQs and timelines
✅ Using the exciting new Audio Overview feature — and even joining the conversation with your voice!
comment in response to
post
During the workshop, I walked participants through a hands-on workflow and demo on how to maximize NotebookLM, including:
✅ Uploading documents, websites, YouTube videos, and more
✅ Summarizing and generating insights instantly
comment in response to
post
Thanks to GDG Mien Trung and the amazing community in Da Nang for the warm energy and insightful questions! 🇻🇳
#AI #AgenticAI #LLM #BuildWithAI #GDG #ArtificialIntelligence #TechCommunity #FutureOfWork #DaNang #OpenAI #VertexAI #LangChain #CrewAI #AutoGen #PydanticAI
comment in response to
post
💡 One big takeaway: AI agents are no longer just tools—they’re becoming collaborators. As autonomy, reasoning, and adaptability evolve, the future belongs to systems that can think, act, and learn alongside us.
comment in response to
post
🧱 Frameworks like LangGraph, CrewAI, OpenAI Agent SDK, and AutoGen that make building agents easier
🔮 Key trends: multi-agent collaboration, event-driven adaptability, ethical AI, and democratization of services
comment in response to
post
🧠 How Large Language Models (LLMs) supercharge agentic workflows
🔄 Emerging AI agent workflows: prompt chaining, routing, parallelization, orchestration, and more
comment in response to
post
From humble beginnings as rule-based systems to today’s autonomous, LLM-powered agents—AI has come a long way. We dove into:
🔍 What makes an AI agent more than automation
⚙️ Different types of agents—reflexive, goal-based, utility-driven, learning & hierarchical